The calanoid copepod Eurytemora affinis from the northern Baltic Sea was exposed to cell-free filtrates of the toxic haptophyte Prymnesium parvum as well as to cell mixtures of P. parvum and Rhodomonas salina. To test the effects of P. parvum exudates and allelopathy on selective grazers, copepods were incubated (1) in increasing concentrations of cell-free filtrates of P. parvum in the presence of good food (R, salina), (2) in 1:1 cell mixtures at 2 cell concentrations of P. parvum and R. salina and (3) in R. salina cell suspension, which was used as a control for good-quality food. P. parvum cultures were grown in nutrient-balanced (+NP) or limited (-N or -P) media to obtain different levels of toxicity. Survival, ingestion, faecal pellet production rates and egg production were measured over 3 d, together with measurements of P. parvum toxicity (hemolytic activity) (HA). Most of the copepods incubated in high-filtrate concentrations died or became severely impaired, although (HA) in filtrates was under the detection limit. Further, the ingestion and faecal pellet production rates were suppressed in the highest filtrate concentrations in nutrient-limited treatments. Higher cell density in cell mixtures resulted in significantly lower faecal pellet production, although survival remained high. Our results show that HA is not a good overall indicator of the total harmful effects of P. parvum on grazers. Besides monospecific P. parvum diets, filtrates and cell mixtures have negative effects on grazers, and these effects are stronger under nutrient-depleted conditions; however, the presence of good-quality food lowers harmful effects for copepods. The negative effects caused either by direct intoxication or by food limitation following from strong allelopathic effects of P. parvum on other components of nano- and microplankton suggest that P. parvum blooms have a realistic potential to be deleterious for copepod secondary production, irrespective of the presence of alternative food sources.